Introduction - If you have any usage issues, please Google them yourself
Programmed based on k-means clustering, for example: randomly select k central point has been calculated for each point to the k center distances, will be classified. The final total distance squared difference between the smallest stops.